Skip to content
Language
Colour scheme

Daily News

What changed in AI today? For the bigger picture, watch the weekly edition.

· Verified sources
The Creative Gap0:00 / 4:48

Get the Daily News by email at 07:00

Free · Unsubscribe anytime

Subscribe to receive the daily AI briefing and/or new explainer videos straight to your inbox.

What would you like to receive in your newsletter?

Research window: 1 November 2026 – 3 November 2026 · Event dates checked separately

Weekly NewsThe week's key events, in context

Runway Integrates Video Tools Directly into Coding Agents

Today highlights new developer plugins for media workflows and an open-weight science model.

Runway Launches Developer Plugins for Claude Code, Codex, and Cursor

On November 2, 2026, Runway launched developer plugins for major AI coding assistants. The tools connect Claude Code, Codex, and Cursor directly to Runway APIs using the Model Context Protocol. Software developers can now invoke generative video features straight from their code editors. The plugin allows an agent to check project credits, choose video models, and call prebuilt pipelines. Developers do not need to store secret API keys inside the plugin. They sign into their developer account once when the agent starts work. When errors occur, the assistant can inspect model logs directly inside the editor.

What it means: This release shows coding agents moving beyond text files to orchestrate complex creative APIs. It cuts integration friction, though it ties team automation closely to one provider's pipeline.

Example: A developer asks an editor agent to script a product page and automatically render short demo clips for each item.

Also worth knowing

Deep Principle Releases Open-Weight Science Model Hydrogen-1.0

On November 2, 2026, Deep Principle released Hydrogen-1.0 under an MIT license. The model contains 753 billion parameters and was post-trained from GLM-5.3. It focuses on chemistry, materials science, and biology. The release aims to give researchers an inspectable assistant for hypothesis generation and laboratory planning.

What it means: Open scientific weights allow academic labs to audit automated discovery tools without relying on proprietary cloud services.

Example: A university chemistry team runs the model locally to draft steps for synthesizing a new polymer.

Anthropic Discloses $17B Revenue and Heavy Losses in Amended S-1

Anthropic filed an amended S-1 registration on November 2, 2026. The filing showed roughly $17 billion in third-quarter revenue, alongside a quarterly net loss of $23.29 billion. It also revealed $895.4 billion in forward purchase commitments and leases, mostly for computing infrastructure.

What it means: The disclosures show the massive financial scale of frontier AI, where record revenues still lag behind heavy infrastructure spending.

Example: An analyst reviews the filing to see how long-term server commitments affect future subscription pricing.

Previous briefings

Powered by anthropic/claude-haiku-5.5

Note

Stays in this browser. Nothing is sent.

Subscribe to the newsletter

Choose Daily News, explainer videos, or both. You can unsubscribe from each separately.

What would you like to receive in your newsletter?

Confirm your subscription by email. No spam; unsubscribe anytime.

Contact

A question, a project or feedback? Send me a message and I will reply by email.

0 / 4000